The Last Five Per Cent: Why Retention Release Deserves a File, Not a Favour

Retention is one of the few commercial tools that survives from first tender to final account. In most UK construction and fit-out contracts it is expressed as a percentage — commonly three to five per cent — withheld from each payment and released in stages. The first tranche usually follows practical completion; the second follows the end of the rectification period, often twelve months later. The mechanism is familiar. The discipline around releasing it is not.

Too often, the final retention is treated as a gesture. A project closes, the team disperses, and someone signs off the last payment because the building looks finished and the contractor has been patient. That is not commercial management. It is a favour dressed up as a process. Retention is a contractual tool, and releasing it should be an evidenced decision, not a reflex.

Question one: What does the defect record actually say?

Before any retention is released, the defect record should be current, specific and signed. A list of snagging items is not the same as a defect schedule. Snagging belongs to completion; defects belong to the rectification period. If the record is vague — 'minor items outstanding' or 'to be confirmed on site' — the release decision is being made on memory rather than evidence.

Ask whether each item has an owner, a date and a status. Ask whether the contractor has closed the items it claims to have closed, and whether the client has inspected that closure. If the answer is 'we think so', the file is not ready. A retention release should be able to point to a dated record that shows either no outstanding defects or an agreed plan for those that remain.

Question two: Is completion evidenced, or merely announced?

Practical completion is a contractual event, not a mood. It should be certified, recorded in writing and accompanied by the documents the contract requires: as-built information, operation and maintenance manuals, test certificates, warranties and any training records. A contractor saying 'we are practically complete' is an invitation to inspect, not a release trigger.

The distinction matters because the first retention tranche is usually tied to practical completion, while the second is tied to the end of the rectification period. If the completion evidence is thin, the release decision inherits that weakness. A client who pays retention against an uncertified completion has little leverage if defects emerge later. The question is not whether the site looks open. It is whether the contract says completion has occurred, and whether the file proves it.

Question three: Has the final account been fixed in time?

Retention release and final account agreement are related but not identical. A final account can be agreed before retention is released; retention can be released before every final account line is settled. The risk is drift. If the final account remains open, the retention becomes a bargaining chip by default, and the release decision gets postponed for reasons that have nothing to do with defects.

Before release, ask whether the final account is agreed in principle, whether any disputed sums are identified, and whether the retention is being withheld against a specific contractual entitlement rather than general caution. Time matters here. Under many standard forms, the employer must release retention within a stated period after the triggering event. Missing that period can expose the client to interest or contractual remedies. A file that shows the account status is stronger than an email that says 'we are still looking at it'.

Question four: Are contractual remedies being used, or replaced by goodwill?

Retention exists to give the client a remedy if the contractor fails to rectify defects. It is not a substitute for managing the contract. If defects are outstanding, the correct response is to follow the contractual mechanism: notify, require rectification, and only then consider whether retention should be withheld or applied. Withholding retention without a contractual basis is not robust; releasing it without checking the basis is equally weak.

The commercial question is whether the client is using the remedy the contract provides. If the answer is that the team is 'keeping the retention as leverage' but has no defect notice, no rectification schedule and no formal position, the leverage is illusory. A contractor can challenge an undocumented withholding. A documented position — even one that leads to release — is more defensible.

Question five: Do you know the difference between practical completion and true close-out?

Practical completion ends the construction phase. Close-out ends the contract. They are not the same, and retention release sits between them. True close-out includes final account agreement, resolution of defects, handover of final documentation, and confirmation that warranties and guarantees are in place.

A client who treats practical completion as close-out will release retention too early or hold it too long. Neither is good. The first weakens the defect remedy; the second creates unnecessary friction and potential interest exposure. The discipline is to map the release trigger to the contract, not to the project team's sense of closure.

A file, not a favour

None of this requires hostility. It requires a file. A retention release decision should be able to answer five simple questions with documents rather than recollections: what defects remain, what completion evidence exists, where the final account stands, what contractual remedy is being exercised, and which close-out stage has actually been reached.

If those answers are not available, the right response is not to release the money and hope. It is to fix the record first. Retention is the last five per cent, but it is not a tip. It is a contractual instrument, and it deserves the same discipline as the first payment.